1. Field of the Invention
This invention relates to an image processing device, and more particularly to an image processing device which extracts character, graphic and photograph regions from image data acquired by reading an original document and performs image processing for the regions and then re-synthesizes them to prepare a document image file.
2. Description of Related Art
Image data read by unit of a scanner or a like device has a great volume and hence is not suitable for storage or for transmission/reception directly as it is. Therefore, a suitable image process such as compression is performed for the image data in accordance with the type of the image. However, in image data acquired by reading an original document, character regions including character images, graphic regions formed from graphic images and photograph regions formed from photographs are included in a mixed state. Therefore, there is a problem that, if compression suitable for the photograph regions is applied to the image data, then this makes it less easy to read the characters although the volume decreases, but if compression suitable for the character regions is applied, then a sufficiently high compression ratio cannot be achieved. Thus, some known image processing device extracts character, graphic and photograph regions separately from one another from image data of an original document and applies different compressions suitable for the individual regions and then re-synthesizes them to prepare a document image file. According to the image processing device just descried, an image file can be prepared with a sufficiently reduced volume while an image quality is maintained (refer to, for example, Japanese Patent Laid-Open No. 2001-169120).
However, depending upon the object of use of the image processing device, various cases are possible. In particular, it is desired to give a preference to the processing speed even if the volume or the picture quality of an image file to be outputted is sacrificed. Or, it is desired to give a preference to the picture quality even if the volume or the processing speed is sacrificed. Or else, it is desired to reduce the volume even if the picture quality or the processing speed is sacrificed. However, an image processing device is not available as yet which can automatically execute a compression process which satisfies any of the demands of users. Also an image processing device is not available as yet with which a user can arbitrarily designate a compression processing method to be applied to each of extracted regions.
Further, in the image processing device described above, if data of a specific region is processed by a default compression method, that is, a compression method suitable for the region, then the image size after the compression process of data of a region whose image size is smaller than a predetermined value sometimes becomes greater than that before the compression process. In order to solve the problem, several methods are available. According to one of the methods, compression processing is performed using a plurality of compression methods and one of resulting data which exhibits a minimum image size is adopted (refer to, for example, Japanese Patent Laid-Open No. Hei 11-215382). According to another one of the methods, compression processing is not performed for a region with which the phenomenon described above occurs (refer to, for example, Japanese Patent Laid-Open No. Hei 9-149248). However, the former method has a problem in that it involves complicated processing because it is necessary to perform compression processing for all regions individually using a plurality of compression methods. Meanwhile, the latter method has a problem in that the compression ratio of the image file as a whole is low because compression processing is not performed for a particular region.